The present invention relates to a network system, a machine allocation device, and a machine allocation method. In particular, the present invention relates to a load distribution technique for a machine (a computer, a computing machinery) where a machine on which an application used by a user operates is allocated from plural machines, and more particularly to a machine allocation technique in which an application on a machine which is arranged in a data center or the like is used from a user's home such as a standard home over a wide area virtual private network (VPN).
In recent years, with an increase in the use of portable video reproduction or music reproduction terminals, or network compatible televisions, in user sides such as a user's home such as a standard home, outdoors, or public facilities, contents such as videos, music, or document files are available by any terminal at any point. Thus, in order to make the contents available by any terminal anywhere, there is a need to install a distribution server that holds and distributes the contents. Usually, in the use in the standard home, a network within the home, that is, a home area network (HAN) is connected with the terminal and the distribution server. In recent years, there is a device that can receive the distribution of the contents from the distribution server within the home from a place other than the home over the Internet. The distribution sever of this type is exemplified by a media distribution server standard “digital living network alliance (DLNA)” and “common internet file system (CIFS)” of a file server. However, when the content distribution server is installed in the user's home as described above, there arise the following problems. That is, an expert knowledge of a network server is required to configure the HAN or the distribution server, thereby making such a configuration difficult. There is a need to increase the number of distribution servers every time a storage capacity for holding the contents becomes short. There is a need to install the distribution server for each service. A user must deal with a failure of the distribution server. In order to solve the above problems, in the future, there is being used a service that is so-called “cloud service” which is distributed to the user terminals over a wide area network such as the Internet in which the distribution server is not installed within the home, but installed within a center (hereinafter also called “data center”) which is operated by a service provider such as a communication carrier or a content provider. When the distribution server is thus installed in the data center, since the service provider configures the distribution server and deals with the failure of the distribution server, the user can easily use the distribution server even if the user has no expert knowledge. Also, because an increase in the storage capacity and service variation can be dealt with by addition of the device on the data center side, the user can use the service upon merely requesting the service. The internet is conceivable as a network for connecting the data center and the home. However, in order to realize the same usability as that of the network HAN within the home of the user, there is a need to configure layer 2-virtual private network (L2-VPN) which is a user dedicated identical internet protocol (IP) domain between the data center and the home. As a technique for configuring such an L2-VPN in the wide area, there are provider backbone bridging (PBB), virtual private LAN service (VPLS), and security architecture for internet protocol (IPsec).
The Ethernet (registered trademark) based wide area network of this type is disclosed in U.S. Pat. No. 7,307,990 B2, “SHARD COMMUNICATIONS NETWORK EMPLOYING VIRTUAL-PRIVATE-NETWORK IDENTIFIERS”.